It would be also great to actually see your data model and the queries you ran, 
also where you used indexes and where not.

Some things that I noticed:

Why do postgres and neo4j have different relationship counts? If the model and 
the data is the same the rel-counts should be the same, or?

In your table Neo4j data size is smaller than postgres but in you conclusion 
you state the opposite?

For your query measurements it would be really helpful to see which queries you 
ran how often and if that's a mean or min, max, or to see the percentiles.
Also if those were graph local queries or graph global queries.

I also agree with Peter on trying with a more recent version of Neo4j as a 
comparison, e.g. 2.1.3 with the streaming transactional endpoint

But you're right Cypher is not fully optimized yet, we're currently working on 
a new query optimizer, but most queries should return much quicker.
Perhaps it also depends on how you used cypher (server over rest (streaming or 
not) vs. embedded)

Ping us if you need any help with that.

Cheers,

Michael

Am 22.08.2014 um 17:08 schrieb Peter Neubauer <[email protected]>:

> Honza, 
> Any chance you could run this with Neo4j 1.9.7 and a recent 2.2 milestone? 
> 1.8.3 is ancient.
> 
> /peter
> 
> (snt)-[:frm]->(phn)
> 
> On Aug 22, 2014 5:03 PM, "Honza AndrÅ¡" <[email protected]> wrote:
> Hi, 
> 
> we run a little benchmark to determine usefullness of relational or graph 
> database for metadata repository in complex data warehouse (based on 
> Teradata). 
> 
> http://mantatools.com/metadata-repository-benchmark-postgresql-vs-neo4j/
> 
> Do you have any other experiences with metadata repository databases? We're 
> running next one this week with Neo4j vs. Titan vs. PostgreSQL to get some 
> more data about different types of graph databases and classic relational 
> ones. 
> 
> 
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "Neo4j" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to [email protected].
> For more options, visit https://groups.google.com/d/optout.
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "Neo4j" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to [email protected].
> For more options, visit https://groups.google.com/d/optout.

-- 
You received this message because you are subscribed to the Google Groups 
"Neo4j" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to